Course content

Basics: Dependability, performance and QoS measures; resources and utilization, failure and repair, system models, analytic modelling, discrete and continuous distributions, basics on stochastic processes, Poisson process, discrete-state - continuous-time Markov models and analysis. Simulation: process-oriented simulation, generation of random variables, problem analysis and model design, DEMOS/SIMULA primitives, output analysis. Performance and traffic: Erlang and Engset blocking models, M/M/1 and Erlang queuing models, Jackson networks. Dependability: reliability, availability and system time in simple redundant systems determined by Markov models, block schemes, fault-trees, structure functions, path and cut sets. Measurements: observation strategies, point and interval estimation, design of experiments.

Learning outcome

An introduction to modelling, analysis, measurements and dimensioning of computing and communication systems with respect to dependability and performance. To use discrete event simulation, probability theory and Markov models as tools in this kind of work. An understanding of stochastic modelling, introduced through examples taken from computing and communication (ICT) systems. (However, detailed knowledge of such systems are not required.) Knowledge of "standard" models and approaches. Extraction of system properties from measurements of real and simulated systems.

Learning methods and activities

Lectures and excersies in analytic evaluation techniques and in simulation. Optional selftests. The course can be held in English if international students attend. If there is a re-sit examination, the examination form may be changed from written to oral.

Recommended previous knowledge

TMA4240/TMA4245 Statistics or similar.

Course materials

Peder J. Emstad, Poul E. Heegaard, Bjarne E. Helvik and Laurent Paquereau, Dependability and performance in information and communication systems. Fundamentals, 278 pages, Inst. for telematikk/TAPIR, 2009.
Graham Birtwisle: DEMOS-A system for Discrete Event Modelling on Simula.

Credit reductions